Airborne minefield detection
project overview
 

This page describes the Airborne minefield detection project.
More details can be found at the official site of the project.






The presentation is structured as follow:
 


Goal

    The main goal of this project is to investigate the potential of airborne remote sensing in the scope of minefield detection. Several sensors will be flown over mine polluted areas. The produced images will be used to detect minefields or minefield indicators.

    In order to demonstrate the feasibility of the approach, two sub goals should be reached:

    1. The detectability of indicators should be demonstrated. Photo interpreters will analyse the images to locate minefield indicators. For this tasks classic tools, such as stereoscopes to get 3D information or simple image processing to enhance the images in order to highlight objects of interest, are used. Some specific image processing tools are also developed.
    2. The possibility to interpret scenes of realistic sizes should be demonstrated. A lot of mine indicators might be quite small. Therefore, high resolution images will be used. For mine polluted regions of size typically encountered in practice, a huge amount of data will be produced. The detection process is a complex task for  which a lot of contextual information must be used. Therefore skilled human photo interpreters must be highly involved and automatic image processing based systems do not seem promising. However, image processing tools that pre-filter the incoming data flux by detecting regions of interests are quite promising. Those regions of interest are then presented to the photo interpreter who will make the final decision.
    Before going to a real mine polluted country, a first phase has been planned to test the whole system in a simulated minefield in Belgium. Phase I has now been completed and validated by the European Commission.

Phase I


Phase II


The partners


Descriptions of sensors used for Phase I

For comments or questions contact: Pascal Druyts or Yann Yvinec.